Multi-omics senescence landmarks facilitate pro-senescence and senolytic sequential cancer therapy

Senescence is recognized as a hallmark of cancer. However, the senescent process can have context-dependent beneficial or detrimental roles in various pathological settings, with distinct pro-tumorigenic or anti-tumorigenic abilities in combination therapy. Here, in this study, we established a senescence signature (SenScore) to estimate tumor senescence status. Multi-omics data, including mRNA expression, miRNA expression, protein expression, somatic mutations, somatic copy number alteration, and clinical data across 33 cancer types, were acquired from The Cancer Genome Atlas. A propensity score matching algorithm was used to balance the effects of potential confounders. We characterized senescence-related molecular features and identified various types of molecular alterations associated with both drug-resistant and drug-sensitive responses to anti-tumor drugs. We exhibited that 108 clinically actionable genes and 75 FDA-approved drugs were significantly biased on different senescence statuses across multiple cancer types. We further validated the most sensitive discovered drugs, roscovitine, pyrimethamine, and LFM.A13 in both bleomycin and H 2 O 2 -induced senescence conditions. Accordingly, the combined application of roscovitine and bleomycin was shown to produce a greater in vivo anti-tumor effect than bleomycin alone. Our study provides an integrated multi-omics landscape of molecular alterations associated with the tumor senescence process. These findings highlight the potential of leveraging senescence-induced vulnerabilities to design sequential pro-senescence and senolytic combination therapies, offering novel avenues for optimizing clinical cancer treatment strategies.